Published April 2012 | Version v1
Journal article

Probability theory and radiation risk evaluated by a mechanical designer-relationship between probability of radiation disorder event and external exposure dose

  • 1. Saitama Univ., Saitama, Saitama (Japan)

Description

The title relationship was calculated with the probability theory based on data of Research Reports (1946-1975) on the atomic bomb disaster by Radiation Effects Research Foundation (RERF), formerly Atomic Bomb Casualty Commission (ABCC), from the aspect of mechanical engineering. The relationship in actual data of A-bomb exposed dose and solid cancer risk for the radiation deterministic disorder event could be assumed to follow the normal distribution using the law of large number, which fitted the probability density function (PDF). When the relationship in non-exposed man was assumed to be the same as above (but shifted to left), the intersection point of the two functions could be regarded to be the risk 1, of which distance to known threshold was assumed to be 6σ (σ: variance). If the threshold was thought to be localized at 3σ from the mean μ, the event probability (EP) was 0.9987= ca. 1: μ=1/2 threshold and σ=1/6. EP could be calculated by integrating PDF: exempli gratia (e.g.), the incidence and EP of leukocyte decrease (threshold>200 mSv), N (625 mSv, 208 mSv), were around 4.24E-2 -0.00160 and 0.00289 -0.274, respectively, by exposure 50 -500 mSv. For stochastic disorders, PDF exhibited the uniform distribution, where the relationship between the dose vs incidence was proportionally linear. By using similar assumptions above, from the incidence (PDF=0.000155/mSv, uniform), EP of stochastic disorders was calculable: e.g., in thyroid cancer, EPs were 0.78 -0.00015 by exposure 5 Sv -1 mSv. As hematological diseases except malignant tumors could be regarded deterministic, similar calculations to above were applied: e.g., the incidence and EP of thyroid disorders, N (6.66 Sv, 2.22 Sv), were around 0.00383 -0.136 and 0.00280 -0.230, respectively, by exposure 0.5 -5 Sv. EP during the latent phase (late effect) could be fitted to the model with Poisson distribution from data of stomatitis in the Reports: EP after n years, Pn (t) = e-λtαtα/n (λ=EP during 1 year); e.g., a solid tumor with the maximum appearance (12.5%) at 10 years (λt) after exposure was predicted to give its sign at 2 years before actual appearance in 11.2% of exposed people. These statistic analyses were thought useful as a reference for the radiation accident like that in Fukushima atomic power plant. (T.T.)

Additional details

Publishing Information

Journal Title
Kikai Sekkei
Journal Volume
56
Journal Issue
4
Journal Page Range
p. 88-97
ISSN
0387-1045